Conventionally, a mobile station executes a reconnection process at the time of handover failure or becoming out of synchronization. Since the reconnection process increases traffic of the networks, it is controlled not to be performed until the last minute of the permissible limit. Specifically, the mobile station sets a predetermined timer value and activates a timer, at the time of detecting the handover failure or out of synchronization. When the timer expires, the mobile station transmits a reconnection request to a connection target to execute the reconnection process. In this case, the timer value to be set is a fixed one, thereby generating a certain non-communication period in any situation. When out of synchronization or the like occurs while users are talking, a pause or silent period continues. This causes users to execute the disconnection process and make a telephone call again. Also, the communication continues even while a pause or silent period is continuing, and the call fee is charged in some cases. Additionally, cells are searched for during the reconnection process, so the power of the mobile station is wasted. Besides, there is a drawback that the frequency band is occupied.
In Patent Document 1, in order to prevent the competition for resources at the time of reconnection process, mobile stations or base transceiver stations collect data that affect the reconnection collisions or competition for resources, so that the reconnection timer value is determined based upon the collected data. The collected data include the quantity and distribution of the connection request, the number of available resources, expected duration of each connection, expected duration of each connection at the current point of time, priority index of the connection request, the expected number of new connection requests, and the like.